Commentary: EHR designers must look to other industries for inspiration

By now, most of us are practiced at the art of distinguishing a well-designed website or smartphone app from a bad one. The elegant ones are easy to understand, responsive, and capable of quickly getting us from point A to point B.

Unfortunately, while the use of electronic health record systems—which are intended to allow physicians and other healthcare providers to efficiently find and access patient medical records—is at an all-time high within hospitals, these systems can't claim any of those positive attributes.

Not only do poorly designed EHR databases create headaches and waste time for providers, they do much worse than that—a recent JAMA study found that difficulty using EHR systems accounted for upward of 557 patient-safety events between 2013 and 2016. We can clearly see that patients are being hurt by bad EHRs. EHRs also contribute to physician burnout; the number of EHRs a single physician needs to use combined with the multitude of designs that all of them have has led to increased stress for healthcare professionals.

This doesn't need to be a call for wholesale EHR overhauls, as much of the problem seems to be rooted in a more straightforward problem: EHR interfaces currently available lack specificity depending on provider specialty and individuality. Within a hospital, patients see many types of physicians—cardiologists, surgeons, endocrinologists, neurologists and others. Because each physician focuses on a different area of care, data from the EHR that each provider desires are different. The neurologist will want the EHR to quickly present the findings from prior neurological exams and is less likely to be interested in an EKG, which the cardiologist will be seeking.

Providers have limited ability to customize the interface for the key health data variables that they are concerned with before the documentation process, which leads to providers spending 50% of their time working with EHRs. For most major EHR vendors, you can select templates or designs from other physicians around the country. However, there is a limit to the level of customization, as updates to these systems may not be supported and as a result, could break any customizations made.

So while the ideal is that these systems should be modeled to present providers with the critical data specific to their individual needs, allowing them to spend more time with patients and less time on computers is currently not feasible.

When designing effective interfaces, the goal is not to conceal data from providers, but to elevate the critical values that an individual provider is concerned with. For example, the daily progress notes and reports from other providers will still be accessible. The goal, after all, is not more fractured EHR databases—it's smarter ones. Building custom interfaces that will be supported for the next 20 years is difficult. Also the interfaces need to be smart enough to transform as the disease states of the patient evolves. This level of intelligence in the interfaces is not possible. For example, in a hospital during an acute myocardial infarction, the interface would be focused on the critical values and treating the immediate disease, but as the patient is doing better and the transition is to chronic-care conditions the interface would change to focus on health maintenance.

This is a pressing challenge; as healthcare becomes more of a team model, the challenges presented by unwieldy EHR systems will only continue to grow. An increasing number of providers are being integrated into the care team, so critical information about clinicians' disease models, hospital workflow models, and knowledge of providers all need to be seamlessly integrated into the interfaces so providers can get their hands on the specific information they need as soon as they need it.

Those of us in the healthcare field should look to other industries for inspiration. Retailers have honed the art of decreasing the friction between a customer and a final purchase, and EHR designers should aim to decrease the friction between providers and relevant patient data.

Vendors are now focusing on workflow to decrease that friction, but re-imagining the interface to change dynamically based on disease models and knowledge of the providers is just on the edge of possibilities. Unfortunately, as things now stand, EHR systems tend to further increase that friction. Still, with the integration of well-designed interfaces, EHRs have the potential to seamlessly empower providers with the data and knowledge needed to transform patient lives and improve overall healthcare delivery.

Dr. Andrew Boyd is an assistant professor of biomedical and health information sciences at the University of Illinois at Chicago. His research focuses on data simplification and data integration, and he regularly educates physicians on how to adopt and implement electronic health records.